Connecticut Regional Haze
State Implementation Plan July 15, 2009
Attachment M
Technical Support Document for 2002 MANE-VU
SIP Modeling Inventories, Version 3
TECHNICAL SUPPORT DOCUMENT FOR 2002 MANE-VU
SIP MODELING INVENTORIES, VERSION 3
Prepared by
E.H. Pechan & Associates, Inc. 3622 Lyckan Parkway, Suite 2005
Durham, NC 27707
November 20, 2006
for the
CONTENTS
TABLES ... vi
FIGURES... viii
ACRONYMS AND ABBREVIATIONS ... ix
CHAPTER I – INTRODUCTION... 1
A. What is the purpose of this TSD? ... 1
B. What are Versions 1, 2, and 3 of the 2002 MANE-VU Inventory?... 2
C. How is this TSD organized? ... 3
CHAPTER II – POINT SOURCES... 4
A. General Methods for all State and Local Agencies ... 4
1. What Data Sources Were Used? ... 4
2. What Quality Assurance Steps Were Performed? ... 5
3. Version 3 Emissions Summary ... 18
B. State-Specific Methods ... 20 1. Connecticut ... 20 2. Delaware ... 21 3. District of Columbia ... 21 4. Maine ... 22 5. Maryland ... 22 6. Massachusetts ... 23 7. New Hampshire ... 23 8. New Jersey ... 24 9. New York... 24
10. Pennsylvania (State, Excluding Allegheny and Philadelphia Counties) ... 25
11. Pennsylvania (Allegheny County, FIPS code 42003)... 27
12. Pennsylvania (Philadelphia County, FIPS code 42101) ... 28
13. Rhode Island ... 28
14. Vermont ... 30
C. What Issues Need to be Addressed in Future Versions? ... 30
CHAPTER III – AREA SOURCES ... 32
A. General Methods for all States... 32
1. What Data Sources Were Used? ... 32
2. What Quality Assurance Steps Were Performed? ... 34
3. Version 3 Revisions ... 53
4. Version 3 Emissions Summary ... 57
B. State-Specific Methods ... 58 1. Connecticut ... 58 2. Delaware ... 59 3. District of Columbia ... 60 4. Maine ... 60 5. Maryland ... 61 6. Massachusetts ... 62 7. New Hampshire ... 63 8. New Jersey ... 65 9. New York... 65
12. Vermont ... 72
C. What Issues Need to be Addressed in Future Versions? ... 73
CHAPTER IV – NONROAD SOURCES ... 75
A. General Methods for all States... 75
1. What Data Sources Were Used? ... 75
2. What Quality Assurance Steps Were Performed? ... 79
3. Version 3 Emission Summaries ... 82
B. State-Specific Methods ... 84 1. Connecticut ... 84 2. Delaware ... 84 3. District of Columbia ... 85 4. Maine ... 86 5. Maryland ... 87 6. Massachusetts ... 88 7. New Hampshire ... 89 8. New Jersey ... 89 9. New York... 90 10. Pennsylvania ... 91 11. Rhode Island ... 92 12. Vermont ... 93
CHAPTER V – ONROAD SOURCES ... 94
A. General Methods for All States... 94
1. Data Sources ... 94
2. What Quality Assurance Steps were Performed? ... 95
3. Version 3 Emission Summaries ... 96
B. State-Specific Methods ... 97 1. Connecticut ... 97 2. Delaware ... 99 3. District of Columbia ... 100 4. Maine ... 102 5. Maryland ... 103 6. Massachusetts ... 105 7. New Hampshire ... 106 8. New Jersey ... 108 9. New York... 110 10. Pennsylvania ... 111 11. Rhode Island ... 113 12. Vermont ... 114
CHAPTER VI – BIOGENIC SOURCES ... 117
A. General Methods for all States... 117
1. What Data Sources Were Used? ... 117
2. Version 3 Emissions Summary ... 118
B. State-Specific Methods ... 118
CHAPTER VII. TEMPORAL, SPECIATION, AND SPATIAL ALLOCATION PROFILES AND PREPARATION OF SMOKE (IDA) AND RPO DATA EXCHANGE PROTOCOL (NIF 3.0) FORMATS ... 119
2. Nonroad Sources... 120
3. Onroad Sources... 120
B. Speciation Profiles ... 121
1. Point and Area Sources... 121
2. Nonroad Sources... 121
3. Onroad Sources... 121
C. Spatial Allocation Profiles ... 122
D. Preparation of SMOKE (IDA) and RPO Data Exchange Protocol (NIF 3.0) Formats .. 122
CHAPTER VIII. METHODS FOR AREAS OUTSIDE OF THE MANE-VU REGION ... 155
CHAPTER IX. REFERENCES ... 157
APPENDIX A - POINT SOURCE INVENTORY, VERSION 3: DATA SOURCES BY SCC, EMISSION TYPE PERIOD, AND POLLUTANT...A-1
APPENDIX B - AREA SOURCE INVENTORY, VERSION 3: DATA SOURCES BY SCC, EMISSION TYPE PERIOD, AND POLLUTANT...B-1
APPENDIX C - NONROAD SOURCE INVENTORY, VERSION 3: FINAL COUNTY, MONTHLY NATIONAL MOBILE INVENTORY MODEL (NMIM) INPUTS TEMPORAL PROFILES FOR POINT AND AREA SOURCES...C-1
TABLES
Table II-1. Description of the Field Names and Descriptions for the SCC Control Device
Ratios Table ... 14 Table II-2. Version 3 2002 MANE-VU Point Source Emissions by State... 19 Table II-3. Connecticut 2002 Point, Version 3: Unique List of Start Date, End Date, and
Emission Types... 20 Table II-4. Delaware 2002 Point, Version 3: Unique List of Start Date, End Date, and Emission
Types... 21 Table II-5. District of Columbia 2002 Point, Version 3: Unique List of Start Date, End Date, and Emission Type ... 21 Table II-6. Maine 2002 Point, Version 3: Unique List of Start Date, End Date, and Emission
Types... 22 Table II-7. Maryland 2002 Point, Version 3: Unique List of Start Date, End Date, and Emission
Types... 22 Table II-8. Massachusetts 2002 Point, Version 3: Unique List of Start Date, End Date, and
Emission Types... 23 Table II-9. New Hampshire 2002 Point, Version 3: Unique List of Start Date, End Date, and
Emission Types... 23 Table II-10. New Jersey 2002 Point, Version 3: Unique List of Start Date, End Date, and
Emission Types... 24 Table II-11. New York 2002 Point, Version 3: Unique List of Start Date, End Date, and
Emission Type ... 25 Table II-12. Pennsylvania 2002 Point, Version 3: Unique List of Start Date, End Date, and
Emission Types... 26 Table II-13. Pennsylvania - Allegheny County 2002 Point, Version 3: Unique List of Start Date,
End Date, and Emission Types ... 27 Table II-14. Pennsylvania - Philadelphia County 2002 Point, Version 3: Unique List of Start
Date, End Date, and Emission Types... 28 Table II-15. Rhode Island 2002 Point, Version 3: Unique List of Start Date, End Date, and
Emission Types... 30 Table II-16. Vermont 2002 Point, Version 3: Unique List of Start Date, End Date, and Emission
Types... 30 Table III-1. Summary of MANE-VU-Sponsored Inventories Included in Version 1 of the Area
Source Consolidated Emissions Inventory ... 41 Table III-2. Area Source Industrial, Commercial/Institutional, and Residential Fossil Fuel
Combustion Uncontrolled Emission Factors for PM10-PRI/FIL, PM25-PRI/FIL, and PM-CON ... 44 Table III-3. SCCs for which PM25-PRI Emissions were Estimated by Applying a Ratio to the
PM10-PRI Emissions in the State inventory ... 49 Table III-4. Revisions to PM25-PRI and PM25-FIL Emissions for Paved and Unpaved Roads
and Construction ... 56 Table III-5. Version 3 2002 MANE-VU Area Source Emissions by State ... 57 Table III-6. Connecticut 2002 Area, Version 3: Unique List of Start Date, End Date, and
Emission Types... 58 Table III-7. Delaware 2002 Area, Version 3: Unique List of Start Date, End Date, and Emission
Table III-8. District of Columbia 2002 Area, Version 3: Unique List of Start Date, End Date,
and Emission Types ... 60
Table III-9. Maine 2002 Area, Version 3: Unique List of Start Date, End Date, and Emission Types... 60
Table III-10. Maryland 2002 Area, Version 3: Unique List of Start Date, End Date, and Emission Types... 61
Table III-11. Massachusetts 2002 Area, Version 3: Unique List of Start Date, End Date, and Emission Types... 62
Table III-12. New Hampshire 2002 Area, Version 3: Unique List of Start Date, End Date, and Emission Types... 64
Table III-13. New Jersey 2002 Area, Version 3: Unique List of Start Date, End Date, and Emission Types... 65
Table III-14. New York 2002 Area, Version 3: Unique List of Start Date, End Date, and Emission Types... 66
Table III-15. Summary of New York's Revisions to Version 3 of MANE-VU's Area Source Inventory ... 68
Table III-16. Pennsylvania 2002 Area, Version 3: Unique List of Start Date, End Date, and Emission Types... 70
Table III-17. Rhode Island 2002 Area, Version 3: Unique List of Start Date, End Date, and Emission Types... 71
Table III-18. Vermont 2002 Area, Version 3: Unique List of Start Date, End Date, and Emission Types... 72
Table IV-1. List of Unique Aircraft, Commercial Marine, and Locomotive SCCs Reported by States in MANE-VU Inventory ... 76
Table IV-2. Summary of Basis for 2002 MANE-VU Aircraft, Commercial Marine, and Locomotive Inventory... 77
Table IV-3. Data Source Code Descriptions... 82
Table IV-4. Annual 2002 Nonroad Sector Emissions by MANE-VU State... 82
Table IV-5. Annual 2002 NONROAD2005 Model Emissions by MANE-VU State ... 83
Table IV-6. Annual 2002 Aircraft, Commercial Marine, and Locomotive Emissions by MANE-VU State... 83
Table IV-7. Delaware NONROAD External Data Files... 85
Table IV-8. Maryland NONROAD External Data Files ... 87
Table V-1. Annual 2002 Onroad Sector Emissions by MANE-VU State ... 97
Table V-2. Connecticut Onroad Data in SMOKE Input Files ... 98
Table V-3. Delaware Onroad Data in SMOKE Input Files ... 100
Table V-4. District of Columbia Onroad Data in SMOKE Input Files ... 101
Table V-5. Maine Onroad Data in SMOKE Input Files ... 102
Table V-6. Maryland Onroad Data in SMOKE Input Files... 104
Table V-7. Massachusetts Onroad Data in SMOKE Input Files ... 105
Table V-8. New Hampshire Onroad Data in SMOKE Input Files ... 107
Table V-9. New Jersey Onroad Data in SMOKE Input Files... 108
Table V-10. New York Onroad Data in SMOKE Input Files... 110
Table V-11. Pennsylvania Onroad Data in SMOKE Input Files ... 112
Table V-12. Rhode Island Onroad Data in SMOKE Input Files ... 113
Table VII-1. Profiles, Cross-references, and Documentation for Model Inputs for Version 3 of
2002 MANE-VU Inventory ... 123
Table VII-2. Point Source Temporal Cross-reference Additions ... 126
Table VII-3. Unknown SCCs in the MANE-VU Point Source Inventory... 130
Table VII-4. Area Source Temporal Cross-Reference Updates ... 131
Table VII-5. Area Source Temporal Cross-Reference Additions... 136
Table VII-6. Area Source Temporal Cross-Reference and Profile Additions for the MANE-VU Inventory ... 141
Table VII-7. Point Source Speciation Profiles Added to Speciation Cross-reference File for CB-IV with PM Mechanism... 151
Table VII-8. Point Source SCCs Lacking Speciation Profile Assignments for CB-IV with PM Mechanism... 152
Table VII-9. Summary of Version 3 Mass Emissions and SMOKE Input Files ... 153
Table VII-10. Unique List of Start Date, End Date, and Emission Type Combinations for Daily Emissions in the MANE-VU 2002 Point and Area Source Inventories, Version 3 ... 154
Table VIII-1. Description of Non-MANE-VU Region Inventory Data Used for MANE-VU BaseB Modeling... 156
Table C-1. MANE-VU County, Monthly NMIM/NONROAD Inputs... C-2 FIGURES Figure 1. MANE-VU 12-Kilometer CMAQ Modeling Domain ... 2
ACRONYMS AND ABBREVIATIONS
ATP Anaerobic Thermal Processor BEIS Biogenic Emissions Inventory System CAA Clean Air Act
CAIR Clean Air Interstate Rule CAMD Clean Air Markets Division CAP criteria air pollutant
CE Control Equipment (NIF 3.0) table CEM Continuous Emissions Monitoring CENRAP Central Regional Air Planning Organization CERR Consolidated Emissions Reporting Rule CMU Carnegie Mellon University CNG compressed natural gas
CO carbon monoxide
CO2 carbon dioxide
EF emission factor
EFIG Emission Factors and Inventory Group EGU electricity generating unit
EI inventory
EM Emission (NIF 3.0) table
EP Emission Process (NIF 3.0) table EPA U.S. Environmental Protection Agency ERP Emission Release Point (NIF 3.0) table ETBE ethyl tertiary butyl ether
ETOH ethanol
ETS Emission Tracking System EU Emission Unit (NIF 3.0) table
FIPS Federal Information Processing Standard FIRE Factor Information and REtrieval Factor GIS geographic information system GSE ground support equipment HAP hazardous air pollutant
HC hydrocarbon
HPMS Highway Performance Monitoring System
ID identification
IDA Inventory Data Analyzer format I/M inspection and maintenance
km kilometer
LAI leaf area indices LEV low emission vehicle LPG liquified petroleum gas
MACT maximum achievable control technology MANE-VU Mid-Atlantic/Northeast Visibility Union
MARAMA Mid-Atlantic Regional Air Management Association MTBE methyl tertiary butyl ether
NEI National Emissions Inventory
NH3 ammonia
NIF NEI Input Format
NMIM National Mobile Inventory Model
NO nitrous oxide
NOx oxides of nitrogen
NYSDEC New York State Department of Environmental Conservation ORIS Office of Regulatory Information Systems
OTC Ozone Transport Commission PAR photosynthetic active radiation PE Emission Period (NIF 3.0) table Pechan E.H. Pechan & Associates, Inc. PFC portable fuel container PM particulate matter PM-CON condensible PM
PM10 particulate matter with an aerodynamic diameter less than or equal to a
nominal 10 micrometers PM10-FIL filterable PM10
PM10-PRI primary PM10
PM2.5 particulate matter with an aerodynamic diameter less than or equal to a
nominal 2.5 micrometers PM25-FIL filterable PM2.5
PM25-PRI primary PM2.5
POTWs public owned treatment works ppm parts per million
psi pounds per square inch QA quality assurance
QAPP Quality Assurance Project Plan RPO Regional Planning Organization RVP Reid vapor pressure
SCC Source Classification Code SPDPRO speed profile
SPDREF speed cross reference SI Site (NIF 3.0) table
SIC Standard Industrial Classification SIP State Implementation Plan S/L State and Local
SMOKE Sparse Matrix Operator Kernel Emissions SO2 sulfur dioxide
TAME tertiary amyl methyl ether TR Transmittal (NIF 3.0) table TSD technical support document U.S. United States
VISTAS Visibility Improvement State and Tribal Association of the Southeast VMT vehicle miles traveled
CHAPTER I – INTRODUCTION
A. What is the purpose of this TSD?
This technical support document (TSD) explains the data sources, methods, and results for preparing Version 3 of the 2002 base year criteria air pollutant (CAP) and ammonia (NH3)
emissions inventories for point, area, onroad, nonroad, and biogenic sources for the Mid-Atlantic/Northeast Visibility Union (MANE-VU) Regional Planning Organization (RPO). The MANE-VU region includes Connecticut, Delaware, the District of Columbia, Maine, Maryland, Massachusetts, New Hampshire, New Jersey, New York, Pennsylvania, Rhode Island, and Vermont. Local air planning agencies include Philadelphia and Allegheny County,
Pennsylvania. The region also includes the Penobscot Tribe of Maine Indian Nation (Tribal code 018) and the St. Regis Band of Mohawk Indians of New York (Tribal code 007). However, these tribal authorities did not provide any data for the 2002 MANE-VU inventory. MANE-VU will use these inventories to support air quality modeling, State Implementation Plan (SIP) development, and implementation activities for the regional haze rule and fine particulate matter (PM) and ozone National Ambient Air Quality Standards (NAAQS).
The inventories and supporting data prepared include the following:
(1) Comprehensive, county-level, mass emissions and modeling inventories for of 2002 emissions for CAPs and NH3 for the State and Local (S/L) agencies included in the
MANE-VU region;
(2) The temporal, speciation, and spatial allocation profiles for the MANE-VU region inventories;
(3) Inventories for wildfires, prescribed burning, and agricultural field burning for the southeastern provinces of Canada; and
(4) Inventories for other RPOs, Canada, and Mexico.
The mass emissions inventory files were prepared in the National Emissions Inventory (NEI) Input Format Version 3.0 (NIF 3.0). The modeling inventory files were prepared in Sparse Matrix Operator Kernel Emissions/Inventory Data Analyzer (SMOKE/IDA) format. Ancillary files (holding spatial, temporal, and speciation profile data) were prepared in SMOKE/IDA compatible format. Figure 1 shows the Models-3 Community Multiscale Air Quality Modeling System (CMAQ) modeling domain for the MANE-VU region.
The inventories include annual emissions for sulfur dioxide (SO2), oxides of nitrogen (NOx),
volatile organic compounds (VOC), carbon monoxide (CO), NH3, and particles with an
aerodynamic diameter less than or equal to a nominal 10 and 2.5 micrometers (i.e., primary PM10
and PM2.5). The inventories included summer day, winter day, and average day emissions.
However, not all agencies included daily emissions in their inventories, and, for the agencies that did, the temporal basis for the daily emissions varied between agencies. The temporal profiles prepared for this project will be used to calculate daily emissions when not available in the inventory files.
Figure 1. MANE-VU 12-Kilometer CMAQ Modeling Domain
B. What are Versions 1, 2, and 3 of the 2002 MANE-VU Inventory?
Work on Version 1 of the 2002 MANE-VU inventory began in April 2004. The consolidated inventory for point, area, onroad, and nonroad sources was prepared by starting with the inventories that S/L agencies submitted to the United States (U.S.) Environmental Protection Agency (EPA) from May through July of 2004 as a requirement of the Consolidated Emissions Reporting Rule (CERR). The EPA’s format and content quality assurance (QA) programs (and other QA checks not included in EPA’s QA software) were run on each inventory to identify format and/or data content issues (EPA, 2004a). E.H. Pechan & Associates, Inc. (Pechan) worked with the MANE-VU S/L agencies and the staff of the Mid-Atlantic Regional Air Management Association (MARAMA) to resolve QA issues and augment the inventories to fill data gaps in accordance with the Quality Assurance Project Plan (QAPP) prepared for this project (MANE-VU, 2004a). MARAMA is the MANE-VU organization’s employees, whereas
MANE-VU is the member S/L agencies plus MARAMA employees. MARAMA is one of three RPOs (in addition to Ozone Transport Commission (OTC) and North East States for Coordinated Air Use Management) supporting the MANE-VU effort.
A draft of the point and area source inventories and summary files were provided for stakeholder review during August 2004. Stakeholder comments were reviewed by the S/L agencies and revisions to the inventory files were made to the files to incorporate stakeholder comments as approved by each S/L agency. The inventories were finalized during December 2004 and the SMOKE input files were prepared and reviewed by the modelers during December 2004 and early January 2005. The final inventory and SMOKE input files were finalized during January 2005.
Work on Version 2 (covering the period from April through September 2005) involved incorporating revisions requested by some S/L agencies on the point, area, and onroad
inventories. Work on Version 3 (covering the period from December 2005 through April 2006) included additional revisions to the point, area, and onroad inventories as requested by some states. Thus, the Version 3 inventory for point, area, and onroad sources were built upon Versions 1 and 2. This work also included development of the biogenics inventory. Version 3 of the nonroad inventory was completely redone due to changes that EPA made to the
NONROAD2005 model.
C. How is this TSD organized?
Chapters II through V of this TSD present the general and State-specific methods and data sources used to develop Version 3 of MANE-VU’s 2002 inventory for point, area, nonroad, and onroad sources. Chapter VI presents the methods, data sources, and model used to develop the biogenics inventory. Chapter VII documents the temporal allocation, speciation, and spatial allocation modeling input files used for Version 3 of MANE-VU’s 2002 inventory for all sectors. Chapter VIII describes the non-MANE-VU region inventory data used for MANE-VU BaseB Modeling. References for the TSD are provided in Chapter IX. Appendices A and B provide the QA Summary Report files prepared during development of the State-specific inventories for point and area sources, respectively. Appendices A and B also provide tables that identify for each S/L agency, the Version 3 data sources, emission type period, pollutant, and the number of counties by source classification code (SCC). For the nonroad inventory, Appendix C provides the final county, monthly National Mobile Inventory Model (NMIM) inputs provided or
confirmed by the States for Reid vapor pressure (RVP), weight percent oxygen, and gasoline sulfur.
CHAPTER II – POINT SOURCES
A. General Methods for all State and Local Agencies 1. What Data Sources Were Used?
Version 3 of the 2002 MANE-VU point source inventory is based primarily on Version 1 with some state-specific revisions incorporated into Versions 2 and 3. Version 1 was developed using the inventories that S/L agencies submitted to EPA from May through July of 2004 as a
requirement of the CERR. All 12 State agencies submitted point source inventories to EPA. In addition, Allegheny and Philadelphia Counties in Pennsylvania each submitted their own point source inventories to EPA. The EPA performed some limited QA review of the S/L inventories to identify format, referential integrity, and duplicate record issues. The EPA revised the inventories to address these issues and made the files available to the S/L agencies on August 6, 2004. These inventory files were used as the starting point for Version 1 of the MANE-VU inventory. These inventory files were obtained from EPA, consolidated into a single data set, subjected to extensive QA review, revised (as approved by the MANE-VU S/L agencies) to address QA issues and to fill data gaps identified while preparing Version 1. Subsequently, the following agencies provided revisions to their point source inventories:
• Version 2 – Connecticut, Delaware, and Maryland • Version 3 – Massachusetts, New York, and Rhode Island
The revisions that these states provided for Versions 2 and 3 are discussed in the “State-Specific Methods” section of this chapter.
In order to track the origin of data, the temporal period of emissions, and to facilitate generation of emission summaries, the following NIF plus fields were added to the Transmittal (TR), Site (SI), Emission Unit (EU), Emission Release Point (ER), Emission Process (EP), Emission Period (PE), Emission (EM), and Control Equipment (CE) tables:
• Data Source Codes: Code Description
S State agency-supplied data.
L Local agency-supplied data to incorporate S/L comments for individual records.
P NH3 emissions from MANE-VU inventory for cement kilns.
AUG-A PM Augmentation: ad-hoc change.
AUG-C PM Augmentation: standard augmentation method.
AUG-O PM Augmentation: set PMxx-FIL = PMxx-PRI for SCCs starting with 10 (external fuel combustion) and 20 (internal fuel combustion). Note: emission factors and particle-size data for estimating condensible
if an agency provided filterable and not primary emissions for these processes. In other words, the primary emissions were assumed to equal the filterable emissions.
AUG-Z PM Augmentation: automated fill-in of zero values where all PM for a particular process is zero.
• Revision Date: This field indicates the month and year during which the last revision was made to a record.
• State Federal Information Processing Standard (FIPS): This field indicates the state FIPS code of the submittal.
• County FIPS: This field indicates the county FIPS code of the submittal. The following NIF plus fields were added to the EM table:
• Emission Ton Value: This field indicates the values of the emissions in tons. This field was used to prepare summaries of emissions on a consistent EU basis.
• Emission Type Period: This field indicates the period of the Emission Type – either ANNUAL or NONANNUAL. This field was used to prepare summaries of annual emissions.
• CAP_HAP: This field identifies records for CAP versus records for hazardous air pollutants (HAPs). For the MANE-VU inventory, the flag is CAP for all records. • Year: This field indicates the year of the data; for this inventory, it is 2002.
Note that the QAPP for Version 1 includes more data source codes than were used in Version 3 of the point source inventory. The data source codes listed above are the codes used in
Version 3. The exception is for Rhode Island, who requested that their Version 2 inventory be replaced with its inventory included in the final 2002 NEI prepared by EPA. Thus, for Rhode Island, it was agreed to maintain the data source codes used in the NEI in Version 3 of the MANE-VU inventory. The data source codes for Rhode Island’s point source inventory are explained under the state-specific section for Rhode Island.
2. What Quality Assurance Steps Were Performed?
A QAPP was prepared and approved by MANE-VU/MARAMA and the EPA Regional Office prior to initiating work on Version 1 of the inventory (MANE-VU, 2004a). This QAPP was followed during preparation of all three versions of the inventory. This section provides an overview of the QA checks completed on each version of the inventory. The QA process for each S/L inventory involved the following steps:
• Conduct QA checks on each S/L inventory;
• Revise the inventory to resolve QA issues as directed by the agency;
• Repeat the QA checks on the revised inventory to verify that the corrections were completed;
• Perform augmentation to correct for missing data; and
• Repeat the QA checks to verify that the augmentation was completed correctly.
a. QA checks for S/L agency inventories
The following discusses the QA diagnoses that were run on the consolidated point source inventory data set. For each S/L agency, a “QA Summary Report” was prepared for each QA check in an Excel Workbook file. The results of each QA check was summarized in a separate spreadsheet and submitted to the S/L agency for review and resolution. The agencies provided corrections to the data in the Excel files or via e-mail and the inventory was updated with the corrections.
i. Continuous Emissions Monitoring (CEM) Analysis
The goal of this analysis was to compare annual NOx and SO2 emissions that were measured with
CEM systems and reported to EPA to the annual NOx and SO2 emissions reported in the S/L
inventories. Facilities report hourly CEM data to EPA for units that are subject to CEM
reporting requirements of the NOx SIP Call rule and Title IV of the Clean Air Act (CAA). Thus,
hourly CEM emissions were summed to the annual level and compared to the annual emissions in the S/L inventories. If the S/L agencies agreed, the CEM hourly emissions would be used to support air quality modeling to accurately reflect the temporal distribution of emissions from CEM units during 2002. Since some of the states require facilities to certify the emissions they report for inclusion in the inventory, the agencies needed proof that the emissions in the CEM inventory compared well with the emissions in the S/L inventory.
The 2002 CEM inventory containing hourly NOx and SO2 emissions and heat input data were
downloaded from the EPA/Clean Air Markets Division’s (CAMD) web site
(www.epa.gov/airmarkets) on July 8, 2004 (CAMD, 2004). The data were provided by quarter and state resulting in 48 separate files for the 12 states in the MANE-VU region. For each state, the hourly emissions were summed to the annual level by facility and EU.
The first stage in the CEM analysis involved preparing a crosswalk file to match facilities and units in the CEM inventory to facilities and units in the S/L inventories. In the CEM inventory, the Office of Regulatory Information Systems (ORIS) identification (ID) code identifies unique facilities and the unit ID identifies unique boilers and internal combustion engines (i.e., turbines and reciprocating engines). In the S/L inventories, the state and county FIPS and state facility ID together identify unique facilities and the EU ID identifies unique boilers or internal combustion engines. However, in some of the S/L inventories, the emissions for multiple EUs were summed and reported under the same EU ID. Thus, an Excel Workbook was sent to the S/L agencies that contained an initial crosswalk with the ORIS ID and unit ID in the CEM inventory matched to the state and county FIPS, state facility ID, and EU ID in the S/L inventory. Agencies were asked to confirm/correct/supplement the information in the crosswalk. The initial crosswalk also contained annual emissions summed from the hourly CEM emissions and flags that indicated if
CEM units were subject to reporting requirements under the NOx SIP Call and/or Title IV of the
CAA. It should be noted that the initial matching of the IDs in both inventories was based on previous crosswalks that had been developed for the 1999 NEI and in-house information compiled by Pechan. The matching at the facility level was nearly complete; however, S/L agency assistance was needed to match most of CEM units to EUs in the S/L inventories.
The crosswalk was updated with corrections to facility and CEM unit-to-EU matches, and with new matches provided by the S/L agencies. The matching of each CEM unit to an EU was still incomplete. Consequently, the comparison of annual emissions was performed at the facility level.
The second stage in the CEM analysis was to prepare an Excel Workbook file for each S/L agency that compared the annual emissions summed from the hourly CEM inventory to the annual emissions reported in the S/L inventory. The file included three spreadsheets that compared annual emissions at the facility level, listed the facilities in the CEM inventory that could not be matched to the facilities in the S/L inventory, and listed the facilities in the S/L inventory identified as an electricity generating unit (EGU) that could not be matched to a
facility in the CEM inventory. The Excel files were sent to the S/L agencies for review. The S/L agencies then indicated if they did or did not want to use the hourly CEM inventory.
The facility-level comparison of CEM to emission inventory NOx and SO2 emissions found that
for some facilities, the annual emissions from the S/L inventory exceeded the CEM annual emissions because the facility in the S/L inventory contained more than just CEM units. This condition was determined to be acceptable. However, S/L agencies were asked to review data for facilities where the CEM emissions were higher than the emissions summed from the S/L inventory. For these cases, CEM emissions may be higher than those reported in a S/L inventory due to methods EPA uses for using artificially high default values to fill in hourly CEM data when not reported or when a CEM unit was not working properly.
After reviewing the comparison of the CEM to S/L inventory emissions, New York and Vermont elected to use the 2002 CEM inventory containing hourly NOx and SO2 emissions for all
facilities. Maryland; New Hampshire; and Allegheny County, Pennsylvania elected to use the 2002 CEM data for some but not all of the facilities within their jurisdiction. The Excel Workbook files containing the comparison of CEM to S/L inventories provides a spreadsheet identifying the facilities for which these S/L agencies elected to use the CEM inventory.
Subsequent to the completion of this analysis, it was determined that the structure of the EPA/CAMD file would not be compatible with the format of the SMOKE input file. The database structure did not affect the annual emissions summed from the hourly CEM emissions used in the comparison to S/L inventory data. For each of the S/L agencies that elected to use the 2002 CEM data, CAMD agreed to provide separate database files for each state with a structure compatible with the SMOKE input file format. Pechan then used the crosswalk to add to the CEM inventory files the state and county FIPS, state facility ID, and EU ID (if the
crosswalk contains a CEM unit to EU match) to the hourly CEM database files provided by CAMD. The modified database was then used to create the SMOKE input files for these states.
Note that Delaware requested that the 2002 CEM inventory for its facilities not be used for regional haze modeling. However, if the consolidated point source inventory prepared under this project is used to support ozone episode modeling, Delaware may consider using the CEM hourly data for the episodes modeled. Therefore, the 2002 CEM inventory was also processed for Delaware’s facilities.
ii. PM Emissions Consistency and Completeness Review
The following consistency checks were performed at the EM table data key level (for annual emissions) to compare PM emissions:
• If a process was associated with a PM emission record, but was missing one or more of the following (as appropriate for the SCC [i.e., condensible PM (PM-CON) is associated with fuel combustion only]): filterable PM10 (PM10-FIL), primary PM10
(PM10-PRI), filterable PM2.5 (PM25-FIL), primary PM2.5 (PM25-PRI), or PM-CON,
the record was flagged for review.
• The following equations were used to determine consistency: PM10-FIL + PM-CON = PM10-PRI
PM25-FIL + PM-CON = PM25-PRI PM-FIL + PM-CON = PM-PRI
• The following comparisons were applied to determine consistency: PM10-PRI >= PM10-FIL PM25-PRI >= PM25-FIL PM10-PRI >= PM-CON PM25-PRI >= PM-CON PM10-FIL >= PM25-FIL PM10-PRI >= PM25-PRI PM-PRI >= PM10-PRI PM-PRI >= PM25-PRI PM-FIL >= PM10-FIL PM-FIL >= PM25-FIL
If the data failed one of these checks it was diagnosed as an error, summarized in an Excel Workbook file, and provided to the S/L agency for corrections. If a S/L agency did not provide corrections to these errors, the errors were corrected or filled in according to the augmentation procedures.
iii. ERP Coordinate Review
Location coordinates for point sources were evaluated using geographic information system (GIS) mapping to determine if the coordinates were within 0.5-kilometers of the boundary of the county in which the source was located. If not, the S/L agency was asked to review the
coordinates and provide corrections to either the coordinates or the state and county FIPS codes. The 0.5-kilometer test resulted in a large number of ERPs for review by the agencies. Therefore, to assist S/L agencies in prioritizing their review of coordinates, ERP records with coordinates located more than 0.5, 1, 2, 3, 5, 7, and 10 or more kilometers from their county boundary, and coordinates that mapped outside of their state boundary were identified. Annual emissions summed to the ERP level were included in the QA Summary Report to identify records with zero emissions for all pollutants and to identify the highest emitting stacks. The QA Summary Report was provided to the S/L agency for review and corrections.
iv. ERP Parameter Review
The EPA’s QA guidance for diagnosing ERP issues for the point source NEI (EPA, 2004b) was applied to identify QA issues in the S/L point source inventories. The QA guidance involved diagnosing the correct assignment of the ERP type (i.e., stack or fugitive), parameters with zero values, parameters not within the range of values specified in the EPA’s QA procedures, and consistency checks (i.e., comparing calculated values against the values reported in the
inventory). In many cases errors were caused by missing or zero values. In other cases, out-of-range errors were caused by unit conversion issues (e.g., stack parameters were in ft, ft/sec, cu ft/sec, or degrees Fahrenheit). The QA issues were summarized in a separate QA Summary Report for each agency and each agency was asked to provide corrections. If an agency did not provide corrections for out-of-range or missing values, the data were corrected or filled in according to the ERP augmentation procedures.
v. Control Device Type and Control Efficiency Data Review The CE codes in the “Primary Device Type Code” and “Secondary Device Type Code” fields were reviewed to identify invalid codes (i.e., codes that did not exist in the NIF 3.0 reference table) and missing codes (e.g., records with a null or uncontrolled code of 000 but with control efficiency data).
QA review of control efficiency data involved diagnosis of two types of errors. First, records were reviewed to identify control efficiency values that were reported as a decimal rather than as a percent value. Records with control efficiencies with decimal values were flagged as a
potential error (although not necessarily an error, since the real control efficiency may be less than 1%).
The second check identified records where 100% control was reported in the CE table, but the emissions in the EM table were greater than zero and the rule effectiveness value in the EM table was null, zero, or 100% (implying 100% control of emissions). Because many agencies did not populate the rule effectiveness field or a default value of zero was assigned, records with null or zero rule effectiveness values were included where the CE was 100% and emissions were greater
than zero. The records that met these criteria were summarized in a QA Summary Report for review and correction, if necessary, by the S/L agency.
vi. Start and End Date Checks
QA review was conducted to identify start date and end date values in the PE and EM tables to confirm consistency with the inventory year in the TR table, and to confirm that the end date reported was greater than the start date reported. This check did not identify any QA issues in the three versions of the inventory.
vii. Annual and Daily Emissions Comparison
The following QA checks were conducted to identify potential errors associated with the incorrect reporting of daily and/or annual emissions:
• Any “DAILY” type record that is greater than its associated “ANNUAL”. A review of the daily vs. annual comparison revealed that in many cases, the daily value was nonzero (but very small), but the annual value was zero. This was generally a result of rounding in a S/L agency’s original emissions database, where annual records were recorded in tons per year to a set number of decimal places, while the corresponding daily records were recorded in pounds per year to a set number of decimal places. The annual record rounds to zero in the original database, while the daily value remains non-zero. A tolerance check reveals the following (comparison in tons):
• Difference Tolerance (daily - annual)> 0 • Difference Tolerance (daily - annual)> .000001 • Difference Tolerance (daily - annual)> .00001 • Difference Tolerance (daily - annual)> .0001 • Difference Tolerance (daily - annual)> .001 • Difference Tolerance (daily - annual)> .01 For Version 1, the affected S/L agencies were as follows:
• Connecticut (09) 11 records • Maine (23) 4 records • Maryland (24) 72 records • New Jersey (34) 2935 records
• Pennsylvania Allegheny County (42003) 17 records • Pennsylvania Philadelphia County (42101) 146 records • Rhode Island (44) 1 record
Rhode Island, Philadelphia, and New Jersey responded that the dailies that were greater than the annuals could be deleted. Maryland determined that they should be kept since the difference values were small. The records for the remaining S/L agencies were kept. This QA issue only
b. Responses from S/L agencies
Each S/L agency reviewed its “QA Summary Report” files and the S/L agency provided direction for correcting QA issues either in the QA Summary Report Excel files or via e-mail. The inventory was then revised to incorporate responses from each agency and the QA checks were run again to verify that the QA issues were addressed. If an agency responded to a QA issue by e-mail, the direction was recorded in the “QA Summary Report” file. The “QA
Summary Report” file for each S/L agency was updated to document QA issues and resolution of issues associated with developing Versions 2 and 3 of the point source inventory. The “QA Summary Report” files for Version 3 are provided with this report in a separate zip file. The files in the zip file are organized in separate folders for each S/L agency. Each folder includes a separate Excel workbook file for the following QA checks if a QA issue existed:
• PM Augmentation QA Summary; • Stack Parameter QA Summary; • Stack Coordinates QA Summary;
• Stack Parameter and Coordinate Augmentation Summary; • CEM Comparisons and Revisions; and
• Control Device/Efficiency Summary.
c. Gap Filling and Augmentation
The following discusses the augmentation procedures that were used to fill in missing data that were not supplied by the S/L agencies. The S/L agencies approved the procedures before they were applied. These procedures were applied after revising the inventory to address QA issues as directed by each S/L agency.
i. MANE-VU-Sponsored Inventories
MANE-VU prepared a 2002 NH3 emissions inventory for cement kilns for SCCs 30500606 and
30500706 located in four MANE-VU states. Maryland chose to add one new facility 24013/0012 (state and county FIPS code/facility ID). New York chose to add the following three sites 36001/4010300016, 36001/4012400001, and 36111/3514800084. Maine and Pennsylvania chose not to add emissions from this inventory. The data for Maryland and New York were added to Version 1. These data were not changed in Versions 2 and 3 of the point source inventory.
ii. PM Augmentation
The PM augmentations process gap-fills missing PM pollutant complements. For example, if a S/L agency provided only PM10-PRI pollutants the PM augmentation process filled in the PM25-PRI pollutants. The steps in the PM augmentation process were as follows:
• Step 2: Development of PM factor ratios based on factors from the Factor
Information and REtrieval (FIRE) Data System, version 6.2, and the PM Calculator (EPA, 2003a; EPA, 2004c);
• Step 3: Implementation of the ratios developed in step 2.; and
• Step 4: Presentation of PM augmentation results to S/L agencies for review and comment.
An Access database (named Reference Tables for PM Augmentation) accompanies this document. This database contains the SCC Control Device Ratio table, the Emission Factors table, and Emission Factors Crosstab table discussed in Step 2. The Emission Factors Crosstab table contains the ratios developed from the Emission Factors table. The Emission Factors table contains detailed information on the emission factors used to develop the ratios. The PM
Calculator ratio table can be provided upon request – it contains all possible combinations for SCC and Control Device types that are available in the PM Calculator. Ratios from the PM calculator were developed using a standard input of 100 TONS of uncontrolled PM-FIL emissions.
1. Initial QA and Remediation of PM Pollutants
S/L agencies were initially presented with files that detailed potential inconsistencies and missing information in their PM pollutant inventory. Inconsistencies in PM pollutants include the following:
• PM-PRI less than PM10-PRI, PM25-PRI, PM10-FIL, PM25-FIL, or PM-CON; • PM-FIL less than PM10-FIL, PM25-FIL;
• PM10-PRI less than PM25-PRI, PM10-FIL, PM25-FIL or PM-CON; • PM10-FIL less than PM25-FIL;
• PM25-PRI less than PM25-FIL or PM-CON;
• The sum of PM10-FIL and PM-CON not equal to PM10-PRI; and • The sum of PM25-FIL and PM-CON not equal to PM25-PRI.
Potential missing information was summarized in a table which detailed the variety of cases provided by each S/L agency. For example, an S/L agency might have provided PM10-FIL and PM25-FIL for some processes, but provided only PM10-FIL for other processes.
S/L agencies were asked to review this information and provide corrections where possible. In general, corrections (or general directions) were provided in the case of the potential
inconsistency issues. An example of a general direction provided by a S/L agency was to remove PM25-FIL where greater than PM10-FIL because the PM10-FIL was (in their particular case) known to be more reliable. In other cases, the agency-provided specific process-level pollutant corrections. If specific direction was not provided by the agency, zero PM pollutants were generally removed, or complements were set equal to the higher number.
2. Development of PM Factor Ratio
The primary deliverable of this step of the process was the development of a table keyed by SCC, primary control device, and secondary control device. This table is called the SCC Control Device Ratios table (see Table II-1). This table was filled according to the following steps:
• Ratios (both condensible and noncondensible) were added from FIRE for SCCs starting with 10* (external fuel combustion) and 20* (internal fuel combustion) where there was a direct match between the provided SCC, and primary and secondary control devices.
• Ratios (non-condensable) were added from the PM Calculator for SCCs starting with 10* and 20* where there was not a direct match between the provided SCC, and primary and secondary control devices. Condensible ratios were added from the PM Calculator based on the uncontrolled SCC for these SCCs. In some cases, it was necessary to map the SCC and control devices to the PM calculator to find a match for the noncondensible ratios. In other cases, it was necessary to map the SCC to FIRE to find a match for condensible ratios.
• For natural gas, process gas, and liquified petroleum gas (LPG) SCCs starting with 10* and 20*, it was assumed (based on FIRE emission factors) that the
PRI/PM10-PRI/PM25-PRI ratio was equal to 1. It was also assumed that the PM-FIL/PM10- FIL /PM25- FIL was equal to 1. Condensible ratios were calculated from uncontrolled FIRE emission factors for these SCCs. In some cases it was necessary to map the SCC to FIRE to find a match for condensible ratios.
• Ratios for SCCs not like 10* and 20* were obtained from the PM Calculator. It was assumed that the condensible component was zero.
Table II-1. Description of the Field Names and Descriptions for the SCC Control Device Ratios Table
Field Name Field Description
PM Calculator A “Yes” in this field indicates that at least some of the information was retrieved from the PM Calculator
FIRE A “Yes” in this field indicates that at least some of the information was retrieved from the Emission Factors table. A “Condensible Ratios” in this field indicates that the condensible ratios factors were retrieved from this table.
Other A field to indicate other sources as necessary.
SCC Source category code from the S/L agency-provided data.
SCC_DESC Description of source category code from the S/L agency-provided data.
maptoSCC This field equals SCC unless the SCC provided was not found in the appropriate source table. In that case, the SCC was mapped using the closest available appropriate mapping choice. maptoSCC_DESC Description of the maptoSCC.
mapSCCNote Any notes related to the mapping of the SCC. A “Yes” in this field indicates that the SCC was mapped.
PD Primary device type from the S/L agency provided data. PD_DESC Description of the primary device (PD).
maptoPD This field equals PD unless the PD provided was not found in the appropriate source table. In that case, the PD was mapped using the closest available appropriate mapping choice.
maptoPD_DESC Description of the maptoPD.
mapPDNote Any notes related to the mapping of the PD. A “Yes” in this field indicates that the PD was mapped. SD Secondary device type from the S/L agency provided data.
SD_DESC Description of the secondary device (SD).
maptoSD This field equals SD unless the SD provided was not found in the appropriate source table. In that case, the SD was mapped using the closest available appropriate mapping choice.
maptoSD_DESC Description of the maptoSD.
mapSDNote Any notes related to the mapping of the SD. A “Yes” in this field indicates that the SD was mapped. PM-FIL/PM10-FIL This field and the following are ratios calculated from emission factors found either in FIRE or the
PM calculator.
PM-FIL/PM25-FIL This field and the following are ratios calculated from emission factors found either in FIRE or the PM calculator.
PM-FIL/PM-PRI This field and the following are ratios calculated from emission factors found either in FIRE or the PM calculator.
PM-PRI/PM10-PRI This field and the following are ratios calculated from emission factors found either in FIRE or the PM calculator.
PM-PRI/PM25-PRI This field and the following are ratios calculated from emission factors found either in FIRE or the PM calculator.
PM10-FIL/PM25-FIL This field and the following are ratios calculated from emission factors found either in FIRE or the PM calculator.
PM10-PRI/PM25-PRI This field and the following are ratios calculated from emission factors found either in FIRE or the PM calculator.
PM-CON/PM10-FIL Condensible ratios were calculate from FIRE if available for 10* and 20* SCCs. If condensible ratios were not found in FIRE for 10* and 20* these ratios were set to zero.
PM-CON/PM10-PRI Condensible ratios were calculate from FIRE if available for 10* and 20* SCCs. If condensible ratios were not found in FIRE for 10* and 20* these ratios were set to zero.
PM-CON/PM25-FIL Condensible ratios were calculate from FIRE if available for 10* and 20* SCCs. If condensible ratios were not found in FIRE for 10* and 20* these ratios were set to zero.
PM-CON/PM25-PRI Condensible ratios were calculate from FIRE if available for 10* and 20* SCCs. If condensible ratios were not found in FIRE for 10* and 20* these ratios were set to zero.
PM-CON/PM-FIL Condensible ratios were calculate from FIRE if available for 10* and 20* SCCs. If condensible ratios were not found in FIRE for 10* and 20* these ratios were set to zero.
PM-CON/PM-PRI Condensible ratios were calculate from FIRE if available for 10* and 20* SCCs. If condensible ratios were not found in FIRE for 10* and 20* these ratios were set to zero.
RPO Specific Note Indicates SCC and control device combinations are in the RPO inventory. Additional Notes Any notes regarding assumptions about ratios.
3. Implementation of the QA Ratios
In order to calculate the additional PM pollutants based on the SCC Control Device ratio table developed in the above step, a crosstab table was created from the EM table based on the following fields: • State FIPS • County FIPS • Tribal Code • EU ID • Process ID • Start Date • End Date • Emission Type • SCC
• Primary Device Type • Secondary Device Type
The primary and secondary device type fields were added based on information from the CE table. If CE information was not available these fields were defaulted to 000
(“UNCONTROLLED”). In the few cases where there was a conflict between the control devices reported for the same process for PM pollutants (e.g., a PM10-PRI is listed as controlled, but PM-PRI did not have control information), the control device type was selected based on the controlled pollutant.
In addition to the fields listed above, the crosstab included the PM emission amounts for the particular process and a field that indicated whether those emissions existed in the inventory. These fields were as follows:
• PM_PRI • PM_FIL • PM10_PRI • PM10_FIL • PM25_PRI • PM25_FIL • PM_CON • PM_PRI_EXISTS • PM_FIL_EXISTS • PM10_PRI_EXISTS • PM10_FIL_EXISTS • PM25_PRI_EXISTS • PM25_FIL_EXISTS • PM_CON_EXISTS
The emission values were in the PM_PRI, PM_FIL, PM10_PRI, PM10_FIL, PM25_PRI, PM25_FIL, PM_CON fields. The _EXISTS field indicated whether the pollutant was provided by the S/L agency. A zero indicated that the pollutant was not provided; a number greater than zero (usually one) indicates that it was provided by the S/L agency.
Prior to the development of this crosstab, the EM table was filled in as much as possible using basic assumptions. For example, if the S/L agency provided zero emissions for some but not all forms of PM for a particular process, it was assumed that all forms of PM for that process were zero and they were filled in accordingly. Since that assumption was that for non 10* and 20* SCCs, the condensible value was zero – that would lead to PM10-FIL = PM10-PRI and PM25-FIL = PM25-PRI and PM-PM25-FIL = PM-PRI. Given that assumption, values for these pollutants were also filled in. After this data insertion, a subset of the crosstab was created. This subset only contained processes that required additional augmentation. The SCC Control Device Type ratio table was based on only those SCC and control device types that required augmentation.
The next step was to fill in the missing information in this crosstab using the information found in the SCC Control Device Ratio table.
In calculating PM complement pollutants, priority was given to calculating –PRI and –CON pollutants. FIL pollutants were only calculated if necessary to calculate other pollutants or if it was a by-product of this calculation.
In augmenting the PM pollutants, the non 10* and 20* SCCs were augmented first, with order given to augmenting based on PM10 where available, PM2.5 where available, and then PM.
Augmenting the PM pollutants for the 10* and 20* SCCs is more complicated, but the basic approach was to augment based on PM10 (FIL or PRI) where available, PM2.5 (FIL or PRI) where
available, and then PM (FIL or PRI) if PM10 or PM2.5 variations were not available. Where both
PM10 (FIL or PRI) and PM2.5 (FIL or PRI) variations were both available, the calculation for
PM-CON was generally driven from the PM10 number and the complements as necessary were
back calculated. Where a PRI emission factor ratio was required and was not available, the FIL emission factor ratio was used.
After completing the calculations, the data was QA checked to ensure that the calculations resulted in consistent values for the PM complement. On a few occasions, the mix of ratio value and the pollutants and values provided by the S/L agency resulted in negative values when FIL was back-calculated. In this case the negative FIL value was set to zero and the PRI value was readjusted. In a few cases the appropriate combination of ratios, SCC, and control efficiencies were not available to calculate the PM10-PRI and PM25-PRI values. In these cases, PM10-PRI and PM25-PRI were set equal. The resultant PM table information was appended to the EM table.
Note: The augmentation procedures resulted in some high condensible ratios that were calculated for some SCC control device type combinations. In most cases, these high
Since the state had already provided the PMxx-PRI records, these PM-CON values were not added.
The data source code field was used to identify records that were added to the inventory to complete the set of PM10-PRI and PM25-PRI emissions.
iii. ERP Coordinates
If an S/L agency did not provide corrections for ERP coordinates that map more than 5 km outside of the county boundary, or provide coordinates for ERP records that did not have any coordinates in the S/L inventory, the following procedures were applied to replace the
coordinates:
• Coordinates for other ERPs at the same facility, if available, that map within the county;
• Coordinates for the centroid of the zip code for a facility if a valid zip code was provided or could be obtained from the agency if it is not valid; or
• County centroid coordinates.
The zip code was taken from the SI NIF 3.0 table. The zip code was compared to a reference table of valid zip codes to verify that it was an active zip code and existed in the state and county reported in the inventory. If a valid zip code for a facility could not be identified, the centroid for the facility’s county was used as a last resort. In some cases, the S/L agency provided confirmation that the S/L coordinates were correct even if the analysis indicated that the
coordinates were outside of the county. These coordinates were not changed. Additionally, all coordinates were converted to latitude/longitude measurements.
iv. ERP Parameters
If valid ERP parameters were not provided by the S/L agency, the ERP augmentation procedures that EPA developed for the 2002 point source NEI were applied to the MANE-VU inventory (EPA, 2004b). It has been determined that the augmentation procedures in this document regarding SCC-specific ERP types and temperatures may be difficult to resolve. When this situation occurs, preference was given to the S/L agency -supplied ERP type and SCC. For example, the procedures do not account for cases where an EU has two processes with one defined as a stack source and the other as a fugitive source. Therefore, the S/L-supplied ERP type was used when this situation occurred. If the ERP type was null, and information was not available from the S/L agency, the stack height information was used as a guide. If stack height information was available, the ERP was treated as a stack, if stack height information was not available, the ERP was treated as a fugitive. An additional modification to the augmentation procedure was also implemented. Since in many cases null values were filled in with zeros by S/L local databases when comparing out-of-range velocities and flows (after it was determined that the stack and diameter information was correct) – null and zero values were treated in the same manner to prevent inappropriate replacement of stack parameter values. Additionally, stack parameter values were rounded to 1 decimal place when comparing with range values (just
for the purposes of comparison) to prevent replacement of S/L parameter values based on negligible decimal differences.
v. Control Device Type and Control Efficiency Data
Control efficiencies that were 100% and rule effectiveness of 100% with non-zero emissions were diagnosed as potential errors and sent to the S/L agencies. Where possible these data were updated with S/L data corrections. Decimal control efficiencies were also diagnosed and sent to the S/L agencies. A decimal control efficiency was usually a sign that a control efficiency had not been entered as a percentage as is required by NIF 3.0. Where possible these data was updated with S/L data corrections.
c. QA Review of Final Inventory
Final QA checks were run on the revised point source inventory data set to ensure that all corrections provided by the S/L agencies were incorporated into the S/L inventories and that there were no remaining QA issues that could be addressed during the duration of the project. The EPA QA program was run on the inventory and the QA output was reviewed to verify that all QA issues that could be addressed were resolved. The QA output file was provided in an Access database along with Version 3 of the inventory.
3. Version 3 Emissions Summary
Table II-2 presents a State-level summary of the annual point source emissions in Version 3 of the 2002 MANE-VU inventory. Note that PM10-PRI and PM25-PRI emissions are included in the inventory for all SCCs for which S/L agencies reported any form of PM, PM10, and/or PM2.5
emissions. If an agency did not report PM10-PRI and/or PM25-PRI but reported PRI, PM-FIL, PM-CON, PM10-PM-FIL, and/or PM25-PM-FIL, the PM augmentation procedures discussed in the TSD were applied to the form of PM emissions supplied by the agency to calculate emissions for the other forms of PM emissions. If an agency reported PM10-PRI and/or PM25-PRI emissions but not PM10-FIL, PM25-FIL, or PM-CON emissions, the agency's inventory was not
augmented to calculate filterable or condensible emissions. Note that PM-CON is associated with only fuel combustion sources.
Table II-2. Version 3 2002 MANE-VU Point Source Emissions by State (Tons/Year) State CO NH3 NOx PM10- FIL PM10- PRI PM25- FIL PM25-
PRI PM-CON SO2 VOC
Connecticut 4,053 12,923 738 1,617 0 1,283 389 15,988 4,907 Delaware 9,766 196 16,345 2,466 4,217 1,919 3,666 1,750 73,744 4,755 District of Columbia 248 4 780 91 161 54 132 68 963 69 Maine 17,005 845 19,939 4,535 7,289 2,567 5,787 2,753 23,711 5,319 Maryland 99,024 305 95,328 3,723 9,029 0 5,054 2,018 290,927 6,184 Massachusetts 21,262 1,463 47,086 2,776 5,852 997 4,161 2,984 101,049 8,263 New Hampshire 2,725 74 9,759 1,180 3,332 786 2,938 2,151 46,560 1,599 New Jersey 12,300 51,593 2,928 6,072 2,543 4,779 3 61,217 14,401 New York 66,427 1,861 118,978 1,808 10,392 1,965 7,080 210 294,729 11,456 Pennsylvania 121,524 1,388 297,379 18,044 40,587 6,038 20,116 5,065 995,175 37,323 Rhode Island 2,234 58 2,764 233 300 117 183 68 2,666 1,928 Vermont 1,078 787 130 304 97 267 2 905 1,097 MANE-VU 357,645 6,194 673,660 38,654 89,150 17,083 55,447 17,462 1,907,634 97,300
B. State-Specific Methods
For each of the MANE-VU states and two local agencies in Pennsylvania, this section identifies the temporal basis of the emissions included in Version 3 and discusses revisions incorporated into Version 3. In addition, this section also discusses the origin of each S/L agency’s emissions included in Version 3. For each agency, a table is provided in Appendix A that lists the data source codes by SCC, emission type period, and pollutant. In addition, an electronic folder is provided for each S/L agency containing the QA Summary Reports prepared during Version 1 and other files documenting revisions included in Versions 2 and 3.
1. Connecticut
Connecticut’s Version 3 point source inventory originates from Version 1 except for the following revisions that Connecticut provided for Version 2 and included in Version 3:
• Changed coordinates for AES Thames, Inc. in New London County to -72.3184, 41.4499 (FIPS code 09011, facility identifier 1544).
• Changed values for Hartford Steam (FIPS code 09003, facility identifier 3471), EU P0250, process 02 for summer daily values as follows: Changed actual throughput from 1934 E6FT3 to 1.934 E6FT3, CO summer daily emissions from 53.185 tons to 0.0532 tons, NOx summer daily emissions from 255.288 tons to 0.1021 tons, and VOC summer daily
emissions from 1.2569 tons to 0.0027 tons.
Table II-3 shows the emission type periods for which Connecticut provided emissions.
Table II-3. Connecticut 2002 Point, Version 3: Unique List of Start Date, End Date, and Emission Types
Emission Type
Period Start Date End Date
Emission Type ANNUAL 20020101 20021231 30 NONANNUAL 20011201 20020228 27 NONANNUAL 20011201 20020228 29 NONANNUAL 20020601 20020831 27 NONANNUAL 20020601 20020831 29
Table A-1 in Appendix A identifies the data sources by SCC, emission type period, and pollutant in the Version 3 point source inventory. This table also shows the number of counties by SCC. Note that some SCC and emission type period combinations are listed more than once because the data source codes are different for more than one SCC and emission type period combination. Connecticut provided the data for CO, NOx, PM10-PRI, SO2, and VOC. Connecticut did not
provide any data for NH3. Emissions for PM10-FIL, PM25-PRI, PM25-FIL, and PM-CON were
calculated from the PM10-PRI emissions provided by Connecticut using the PM augmentation procedures.
2. Delaware
Delaware’s Version 3 point source inventory originates from Version 1 except for some updates to ORIS Boiler IDs in the EU table that were incorporated into Version 2 and included in
Version 3. Table II-4 shows the emission type periods for which Delaware provided emissions.
Table II-4. Delaware 2002 Point, Version 3: Unique List of Start Date, End Date, and Emission Types
Emission Type
Period Start Date End Date
Emission Type
ANNUAL 20020101 20021231 30 NONANNUAL 20020601 20020831 29
Table A-2 in Appendix A identifies the data sources by SCC, emission type period, and pollutant in the Version 3 point source inventory. This table also shows the number of counties by SCC. Note that some SCC and emission type period combinations are listed more than once because the data source codes are different for more than one SCC and emission type period combination. Delaware provided the data for CO, NH3, NOx, SO2, and VOC. Delaware also provided much of
the PM emissions data but in some cases the PM augmentation procedures were applied to the PM data provided by Delaware to calculate emissions for other forms of PM (e.g., to estimate PM10-PRI from PM10-FIL, PRI from FIL, PM10-PRI and PM10-FIL from PM25-PRI and PM25-FIL).
3. District of Columbia
The District of Columbia’s Version 3 point source inventory originates from Version 1. Table II-5 shows the emission type period for which the District of Columbia provided emissions.
Table II-5. District of Columbia 2002 Point, Version 3: Unique List of Start Date, End Date, and Emission Type
Emission Type
Period Start Date End Date
Emission Type
ANNUAL 20020101 20021231 30
Table A-3 in Appendix A identifies the data sources by SCC, emission type period, and pollutant in the Version 3 point source inventory. This table also shows the number of counties by SCC. The District of Columbia provided the data for CO, NH3, NOx, SO2, and VOC. The District of
Columbia provided at least one form of PM emissions and the PM augmentation procedures were applied to the emissions provided by the District of Columbia to calculate emissions for the other forms of PM.
4. Maine
Maine’s Version 3 point source inventory originates from Version 1. Table II-6 shows the emission type periods for which Maine provided emissions.
Table II-6. Maine 2002 Point, Version 3:
Unique List of Start Date, End Date, and Emission Types
Emission Type
Period Start Date End Date
Emission Type
ANNUAL 20020101 20021231 30 NONANNUAL 20020601 20020831 29
Table A-4 in Appendix A identifies the data sources by SCC, emission type period, and pollutant in the Version 3 point source inventory. This table also shows the number of counties by SCC. Note that some SCC and emission type period combinations are listed more than once because the data source codes are different for more than one SCC and emission type period combination. Maine provided the emissions data for CO, NH3, NOx, SO2, and VOC. Maine provided
PM10-FIL and/or PM25-PM10-FIL emissions data and the PM augmentation procedures were applied to the emissions that Maine provided to calculate emissions for the other forms of PM.
5. Maryland
Maryland’s Version 3 point source inventory originates from Version 1 except for some updates to ORIS Boiler IDs in the EU table that were incorporated into Version 2 and included in
Version 3. Table II-7 shows the emission type periods for which Maryland provided emissions.
Table II-7. Maryland 2002 Point, Version 3: Unique List of Start Date, End Date, and Emission Types
Emission Type
Period Start Date End Date
Emission Type ANNUAL 20020101 20021231 30 ANNUAL 20040101 20041231 30 NONANNUAL 20020101 20021231 29 NONANNUAL 20020501 20020930 29 NONANNUAL 20040101 20041231 29
Table A-5 in Appendix A identifies the data sources by SCC, emission type period, and pollutant in the Version 3 point source inventory. This table also shows the number of counties by SCC. Note that some SCC and emission type period combinations are listed more than once because the data source codes are different for more than one SCC and emission type period combination. Maryland provided the emissions data for CO, NH3, NOx, SO2, VOC, PM10-PRI, and PM-PRI.
The PM augmentation procedures were applied to the PM10-PRI emissions that Maryland provided to calculate emissions for the other forms of PM. Maryland provided NH3 emissions
for its point sources except for one new facility (state and county FIPS code 24013, facility ID 0012, SCC 30500622, data source code P) for which it used NH3 emissions for four EUs
6. Massachusetts
Massachusetts’ Version 3 point source inventory originates from Version 1 except for the some stack parameter revisions that Massachusetts provided and were incorporated into Version 3. For Version 3, Massachusetts provided revisions to stack parameters in the ERP table for six EUs at three facilities. The revisions are listed in the Excel file named “MA Revisions to MANEVU V3 Point EI_040706.xls”. Table II-8 shows the emission type periods for which Massachusetts provided emissions.
Table II-8. Massachusetts 2002 Point, Version 3: Unique List of Start Date, End Date, and Emission Types
Emission Type
Period Start Date End Date
Emission Type
ANNUAL 20020101 20021231 30 ANNUAL 20030101 20031231 30
Table A-6 in Appendix A identifies the data sources by SCC, emission type period, and pollutant in the Version 3 point source inventory. This table also shows the number of counties by SCC. Note that some SCC and emission type period combinations are listed more than once because the data source codes are different for more than one SCC and emission type period combination. Massachusetts provided the emissions data for CO, NH3, NOx, SO2, and VOC. Massachusetts
provided PM-FIL, PM10-FIL, and/or PM25-FIL emissions data and the PM augmentation procedures were applied to the emissions that Massachusetts provided to calculate emissions for the other forms of PM.
7. New Hampshire
New Hampshire’s Version 3 point source inventory originates from Version 1.
Table II-9 shows the emission type periods for which New Hampshire provided emissions.
Table II-9. New Hampshire 2002 Point, Version 3: Unique List of Start Date, End Date, and Emission Types
Emission Type
Period Start Date End Date
Emission Type
ANNUAL 20020101 20021231 30 NONANNUAL 20020601 20020831 29
Table A-7 in Appendix A identifies the data sources by SCC, emission type period, and pollutant in the Version 3 point source inventory. This table also shows the number of counties by SCC. Note that some SCC and emission type period combinations are listed more than once because the data source codes are different for more than one SCC and emission type period combination. New Hampshire provided the emissions data for CO, NH3, NOx, SO2, and VOC. New
Hampshire provided PM-FIL, PM10-FIL, and/or PM25-FIL emissions data and the PM augmentation procedures were applied to the emissions that New Hampshire provided to
